Mean seed weights for four soybean cultivars grown in 1973 and 1974 in cylindrical open-top field chambers that provided carbon-filtered air were significantly greater than for plants grown in nonfiltered air in chambers or in conventional plots without chambers. Using this seed, an experiment was designed to answer three questions: does air quality influence seed yields from subsequent plants; do seed size differences, possibly induced by air pollutants, influence subsequent seed yields; is there a yield advantage from planting large seed vs. an original lot of seed. This experiment was designed as a split-split plot with six replications. The paper discusses test methods and results of the experiments.